Arnheim, Gestalt and Art A Psychological Theory.pdf

Ian Verstegen
Arnheim, Gestalt and Art
A Psychological Theory
SpringerWienNewYork
This is the pdf of my 2005 book.
In placing it on the internet, I
have taken the opportunity to
correct some careless errors.
The pagination remains the same.
I.V. 21.6.2017

Ian Verstegen
The University of Georgia Studies Abroad Program
Cortona, Italy
This work is subject to copyright.
All rights are reserved, whether the whole or part of the material is concerned, specifi-
cally those of translation, reprinting, re-use of illustrations, broadcasting, reproduction
by photocopying machines or similar means, and storage in data banks.
© 2005 Springer-Verlag Wien · Printed in Austria
SpringerWienNewYork is a part of Springer Science + Business Media
springeronline.com
Product Liability: The publisher can give no guarantee for all the information contai-
ned in this book. The use of registered names, trademarks, etc. in this publication does
not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
Typesetting: Camera ready by author
Printing: Börsedruck Ges.m.b.H., 1230 Wien, Österreich
Printed on acid-free and chlorine-free bleached paper
SPIN: 11553045
Library of Congress Control Number: 2005933044
ISBN 10 3-211-28864-3 SpringerWienNewYork
ISBN 13 978-3-211-28864-1 SpringerWienNewYork

v
PREFACE
When one hears the words, the ‘psychology of art,’ one is likely to think
of the name of Rudolf Arnheim. Because of his great productivity we
have been fortunate to hear the latest words of wisdom from this remar-
kable nonegenarian, almost up to the present day. But while Arnheim the
personality is always intriguing, his system risks being left behind. Alt-
hough Arnheim has remained in remarkable contact with younger scho-
lars around the world, his ideas have risked alienation from their basic
gestalt basis. This book is a presentation of the whole unified Arnheim
through the lens of a living, breathing Gestalt psychology.
Arnheim’s two complementary works, Art and Visual Percep-
tion (1954/1974) and The Power of the Center (1982/1988) will surely
hold their own in visual art theory for some time to come. But Arnheim,
himself, never attempted to provide a general psychology of art. Nor, it
seems, did he presume he ought to. In fact, he once wrote that the book
by Hans and Shulamaith Kreitler “may well claim to have established
the psychology of arts as a discipline” (1973, p. 647). As much as that
may have been true then, it is much less true now. Arnheim has by now
written on every subject of the psychology of art and a general approach
may be said to exist, if not in one place. This work is an attempt to bring
into a single coherent statement this theory.
I began to discern in Arnheim’s a unified approach centered on
the idea of perceptual dynamics. The Power of the Center (1982/1988)
raised new problems of theoretical exposition, and suggested that its
compositional scheme was the key to this unified approach. Indeed, in
The Dynamics of Architectural Form (1977) Arnheim suggested that “I
have come increasingly to believe that the dynamics of shape, color, and
movement is the decisive, although the least explored, factor of sensory
expession” (p. 7); this, only three years before the book on composition.
This suggested a central model based on ‘the dynamics of architecture,’
‘shape,’ etc.
It gave me the suspicion that Art and Visual Perception (1974)
could then be abstracted into such a form. This would then free the pre-
sentation of general principles (the chapters ‘Dynamics’ and ‘Expressi-
on’) and developmental aspects (the chapter ‘Growth), which could then
have their own separate presentation. I began describing my scheme to
Arnheim and he reacted with interest, and surprise. It “is like a dam
break in the hydraulic system of my work, which for most of my purpo-
ses is a system only because I look at items of my work one piece at a a
time, whereas you are presenting it as a whole” (Arnheim, 1992b). As I

vi
asked about the holes that had appeared with the scheme I had commit-
ted myself to, Arnheim at the same time was busy elucidating what he
called the ‘keystones,’ the elementary ground concepts, of his theory that
had been left unstated. The result is what I like to believe to be two high-
ly complementary ventures.
This book began as the labor of an overambitious youngster, in-
spired by the gentle words of a retired sage-like figure in Ann Arbor,
Michigan. I am very grateful to Prof. Arnheim for the generosity he sho-
wed me then, and over the ensuing years, providing me with a priceless
experience of mentoring.
I brought an early draft of the book to maturity under the encou-
ragment of Tiziano Agostini, Howard Gruber, Kendall Walton and
Wolfgang Wildgen, as well as Alan Gilchrist and John Ceraso, the only
two teachers of psychology I have ever had. A good push to get restarted
was given by Lucia Pizzo Russo, Joseph Glicksohn, Michael Kubovy
and Walter Ehrenstein. I wish to thank them all but especially Tiziano
Agostini and Wolfgang Wildgen for their steadfast support over many
years. To my beloved wife Louise, who has cheerfully come to accept
the presence of Arnheim in our lives, I dedicate this book with thanks.
Cortona, Italy Ian Verstegen

vii
TABLE OF CONTENTS
Introduction Why Arnheim? Why Gestalt? 1
Chapter 1 Arnheim and Different Approaches to the
Psychology of Art 8
Chapter 2 Expression, Symbolism and Visual Thinking 19
Chapter 3 The Senses, Perceptual Objects and their
Dynamics 25
Chapter 4 The Notorious Gestalt Brain Model 37
Chapter 5 The Dynamics of Pictorial, Sculptural and
Architectural Form 45
Chapter 6 The Dynamics of Pantomimic Form 57
Chapter 7 The Dynamics of Musical Form 71
Chapter 8 The Dynamics of Poetic Form 83
Chapter 9 Art in Comparative Perspective:
Children, Adults, Cultures 97
Chapter 10 The Dynamics of Microgenetic Artistic
Development 105
Chapter 11 Individual Artistic Development 115
Chapter 12 The Dynamics of Differential and
Psychopathological Artistic Form 125
Chapter 13 The Dynamics of Cultural-Perceptual Form 135
Chapter 14 Objective Percepts, Objective Values 141
Appendix The Life of a Psychologist of Art 151

1
INTRODUCTION
WHY ARNHEIM? WHY GESTALT?
As we address Arnheim and his writings today, we are faced with
the problem of the exposition of his theories, which has never been
attempted, and the other problem of their defense, which is really
not possible without some sense of the former problem. The very
structure of this book is an implicit ordering of Arnheim's thought.
It attempts to order this thought, however, in terms of basic psy-
chology. That is, it seeks to work its way down, ever more specifi-
cally, from problems of general psychology, the problem of ex-
pression central to perception of works of art, to individual sense
modalities that explain them.
It is impossible, however, to rest content with a mere expo-
sition of Arnheim's positions. We are inevitably led to complete our
understanding of his theories with the Gestalt psychology of his
teachers Max Wertheimer and Wolfgang Köhler. But once this has
begun we have to take account of the Gestalt psychology contem-
porary to Arnheim and ultimately that developing up until yester-
day. I find it impossible to discuss Arnheim's views of perception
without thinking of contemporary developments in perception,
especially as proposed by European scientists working in the gestalt
tradition.
This is not only true because one familiar with the work
naturally sees the relationship but also because many dismiss Arn-
heim simply because they regard Gestalt psychology to be wrong,
outdated or irrelevant. I will amply show, in the following chapters,
both Arnheim's synthetic position and the ways it is bolstered and
amplified in contemporary gestalt-style theory. For the time being,
however, it is worthwhile to sketch the attraction of the gestalt ideal
in the first place, one to which Arnheim contributed, but to which
he was also attracted.
It is tempting to ascribe the gestalt position of an ‘analytic
holism’ or else an ‘organic materialism’ to a unique intersection of
ideas fermenting in the Weimar Republic in Germany with Roman-
tic ideas reaching back to Goethe and beyond to Spinoza. Writing
of his teacher, Max Wertheimer, Arnheim (1986) wrote how “Spi-
nozistic was the notion that order and wisdom are not imposed
upon nature but are inherent in nature itself; of great interest also
was Spinoza's idea that mental and physical existence are aspects of
one and the same reality and therefore reflections of each other”
(p. 37).
But the most exciting thing to Arnheim was the contempo-
rary promise in the sciences of his day. According to Gestalt
thinking, the world and the human mind both share principles of

ordering. It is not a matter of imposing order on nature or escaping
in our minds an irrational outer world, rather, the ways our minds
work is precisely due to the principles that order nature. This
thinking is evident in the work on ‘physical gestalts’ by Arnheim's
other teacher Wolfgang Köhler (1920; Arnheim, 1998).
More specifically, Gestaltists hold to a variety of doctrines
that are appealing, and for which they have consistently found ex-
perimental support. There is obviously the famous gestalt approach
to perception, about which much more will be said. The incorrect
epithet ‘the whole is more than the sum of its parts’ has been ap-
plied to this thinking but the real doctrine is that we perceive the
world as ordered, clear-cut and meaningful. This has wide ranging
ramifications for Gestaltists who have used ideas of perceptual or-
ganization well beyond perception.
Much more than mere perception, an image of humanity
attaches to ordered perception. We perceive the bounty afforded by
some things and the lack missing in others. The need felt by a
helpless child is a command to help. Even in those famous cases
which Arnheim's colleague Solomon Asch (1952) investigated, in
which a group gangs up on an individual and tells them that an
obviously longer line is actually shorter, that person reacts ration-
ally, trying hard to reconcile their basic trust in interpersonal com-
munication with the facts before their eyes. Unfortunately, as so
many gestalt doctrines, this has been interpreted beyond recogni-
tion as evidence that group pressure can make us do practically
anything (Friend, Rafferty & Bramel, 1990).
There is a strong tendency in Gestalt psychology to hold
back at the seeming irrationality of findings, whether they be in
Freudian rationalization, conformity studies or investigations of the
effects of needs and desires on perception, and think through the
epistemological consequences of the ‘naive relativism’ seemingly
evident in the results. It is too easy to invoke an irrationalist model
of human motivation until we consider what the consequences of
our own status as scientists, subject to the same foibles. When we
ignore these consequences we commit what the philosopher
Maurice Mandelbaum, in an appreciation of Arnheim’s teacher
Wolfgang Köhler, called ‘the self-excepting fallacy’ (Mandelbaum,
1984; Verstegen, 2000b). In fact, we soon discover that such irra-
tionalism is not as rampant as we think and it is much more inter-
esting and representative of the real world to consider the balance
of objective input and personal motivations that propel us.
There is no strong gestalt movement today in social psy-
chology but numerous studies throughout the field show a consis-
tency with gestalt interests and give further impetus to the image of
humanity propounded by the Gestaltists. The brilliant work on
altruism and the environment as a cause for motivation by Michael
2

and Lise Wallach as well as Henri Zukier is a case in point (Wallach
& Wallach, 1983; Wallach & Wallach, 1990; Zukier, 1982). These
works confirm in a larger sense what I wish to confirm for Arnheim
and perception. The gestalt school has a necessary scientific mes-
sage to import that uniquely solves many important problems.
Unfortunately, as in the case of social psychology, the aims
of Gestalt psychology have been misunderstood in perception. The
most common response is ignorance. The hopeful starts of the ge-
stalt school were not quantifiable, it is argued, to make a lasting
impression on the field, which had to wait for the cognitive revolu-
tion. A more positive interpretation has been offered by David
Murray, in his recent book, Gestalt Psychology and the Cognitive
Revolution (1995). He argues that many principles championed by
the cognitive revolution, including cognitive schemas, organization
and prototypes, were first discovered by gestalt psychologists.
While often true, this interpretation fails to account for the
lack of recognition of Gestaltism by the pioneers of Cognitive Psy-
chology. In fact, the scientism or positivism of Cognitive psychol-
ogy set it apart from the aims of Gestalt psychology, and this very
quality has more in common with the spirit of Behaviorism that
Cognitivism is universally considered to have replaced. More skepti-
cal interpretations of Cognitivism have emerged from the Gestalt
camp (Vicario, 1978; Henle, 1990), not least from Arnheim. The
need for control, prediction and immediate operationalization of
contemporary psychology is alien to the aims of Gestalt psychol-
ogy.
Now is probably the first time that an adequate defense of
Gestaltism has been able to emerge. After the sometimes-
curmudgeonly defenses of Gestalt positions by Arnheim and oth-
ers, a new generation with a new objectivity has emerged to hon-
estly examine some of the main precepts and assumptions of psy-
chology and its theoretical underpinnings. Foremost have been
anti-representationalist philosophers of mind who look to phe-
nomenologists (especially existential, like Heidegger and Merleau-
Ponty), from which the advantages of the allied Gestalt viewpoint
naturally emerge (Petitot, Roy, Pachoud & Varela, 1998).
Unfortunately, psychological thinking on art has bifurcated
into a technical and often unenlightening perceptualism (Solso,
1994; Zeki, 1999) and a speculative Freudianism. Perceptual psy-
chology’s inability to address meanings is nowhere clearer than in
its discussions of visual art. Normally, little more is accomplished
than pointing out perceptual mechanisms or illusions in actual
works of art. There is a new interest in looking at questions the way
Arnheim proposed, and a new conviction that he and Gestalt psy-
chology were on the right track. To this conviction this book is
addressed.
3

Its structure is basically divided into three parts: founda-
tional principles, these principles applied to the various arts, and
then the developmental aspect of art. Of foundational principles,
the next chapter discusses “Arnheim and Different Approaches to
the Psychology of Art.” Regarding the desirability of an Arn-
heimian psychology of art to be dependent on our definition of
psychology, an attempt is made to first define this in reasonable
terms, and see where Arnheim fits in. After discussing the virtues of
a perceptual approach to art, a distinction is made between ‘Cogni-
tive Nativism’ and ‘Cognitive Inferentialism,’ the two dominant
schools of perceptual research today. Introducing the Gestalt alter-
native, I face the question of the degree to which we can say that a
Gestalt school still survives. Then, after addressing the tenability of
its central principles of Relational Determination and Simplicity, I
sketch the work of other researchers whose Gestalt works on the arts
complement those of Arnheim.
Chapter 2 goes on to address concepts that are important
for all psychology, but particularly important for the psychology
of art, namely, the concepts of “Expression, Symbolism and Visual
Thinking.” Beginning with the idea of emotion as a functional
relationship between subject and object, we first must grant a phe-
nomenal reality to expression. Then, a psychology of structural
similarity (isomorphism) underlying expression, metaphor and
symbolism is possible. Metaphor arises because of differences in
juxtaposed levels of abstraction and symbolism. The abilities to
recognize the commonality of genera is abstraction. The gestalt
doctrine of singularity (prägnanz) accounts for abstracted
saliences in perception. On the other hand, when these
singularities are in-stead sensibly compared and manipulated, on
their way to a new order, productive – or as Arnheim (1969) calls
it, visual – thinking occurs.
Chapter 3, “The Senses, Perceptual Objects and their Dy-
namics,” addresses the fundamental differences between seeing and
hearing as a foundation for the arts expressed in those modalities.
Nevertheless, in spite of differences, general gestalt principles of
organization – Wertheimer’s famous laws – can be applied gener-
ally. Arnheim in The Power of the Center has proposed a simple
system of perceptual centers generating dynamics between them,
and bounded by various formats, a surprisingly powerful system.
Arnheim’s theory of composition, then, becomes a more special-
ized account of perceptual dynamics. This is the basis for the rest
of the book.
First, however, Chapter 4 addresses the “Notorious Gestalt
Brain Model,” the model proposed by Wolfgang Köhler that has
been universally dismissed and with it the entire Gestalt school.
While criticizing the particular way in which it was presented, the
4

overall spirit is defended and much confirmation for it is found in
current work on brain modeling and dynamics. I conclude that
such an approach cannot be done away with, for its monistic spirit
drives the things that is most attractive about Gestaltism, its monistic
naturalism.
Chapter 5, “The Dynamics of Pictorial, Sculptural and Ar-
chitectural Form,” begins the second part of the book, the discus-
sion of individual media. These are not concrete media, like
‘painting’ or ‘cast-bronze sculpture,’ but abstracted modalities of
(1) envisagement, sculpture, architecture and more. This chapter
includes the three visual modalities. Pictures contain shapes, which
are perceptual centers. These develop meaning within the confines
of the format, the frame, page or whatever. The object can be an
object in its own right when the format becomes the environment
around it. Sculptures are centers containing sub-centers but the
format is instead the virtual axis created by the vertical. Buildings,
finally, are complicated by the fact that they (and their sub-centers)
are inhabited, creating a constant interaction between inside and
outside, plan and flow, elevation and outside aspect. The building
itself, as a center, then interacts with the landscape or urban fabric.
Chapter 6, “The Dynamics of Pantomimic Form,” dis-
cusses the psychological principles underlying visual action. ‘Pan-
tomimic’ is intended to capture both real (‘absolute’) movement,
in which a single body is observed in continuous space, and ‘ed-
ited’ movement, made possible by the technology of film, video
and animation. Of absolute movement, we have dance and acting,
and both interact with the format of the stage. Similarly, edited
action has the frame of the film to work with.
Chapter 7, “The Dynamics of Musical Form,” applies
Arnheim’s ideas to music. Tones are centers that derive meaning
from the kind of scale they interact with. In the western diatonic
scale the major and minor modes of the scale are the clearest means
to work with. The arbitrary tonic chosen then chains the tones to
the recurrent ordering of the scale. The tone moves up this scale
and down it, as a coming-toward and a moving-away-from, with
attached notions of striving and rest. Harmony, instead, is a ‘vertical
structure’ that similarly moves with the composition. Together with
the meter, phrase dynamics are developed, and larger compositional
orderings.
Chapter 8, “The Dynamics of Poetic Form,” applies the
dynamics of centers to language. Language can be more (me-
mento) or less (message) poetic. It has numerous layers, the visual,
the tonal, the syntactic and semantic. Verbal dynamics mitigate
between stable centers of nouns and pronouns. Against the back-
ground of a version of the Interaction Theory of Metaphor, poetic
meanings develop in complex interaction of juxtaposed mental
5

images that draw out new meanings. Larger structures emerge too
with degrees of centeredness and diffuseness, helping determine the
style of the work.
The third part of the book treating development begins with
Chapter 9, “Art in Comparative Perspective: Children, Adults, Cul-
tures.” This chapter critically introduces the place of development
within the gestalt tradition, challenging the view that Gestalt theory
cannot model such phenomena. Reflecting on a theory of devel-
opment as perceptual differentiation, the gestalt view is distin-
guished from Werner’s and Piaget’s related approaches. More par-
ticularly, Arnheim’s inspiration in the works of Gustaf Britsch is
discussed. A reconstruction of Arnheim’s general theory of artistic
development is undertaken to affirm its generality but not over-
specification of the path of growth.
This leads directly to the discussion of short-term develop-
ment, Chapter 10, “The Dynamics of Microgenetic Artistic Devel-
opment.” Various tasks like artistic problem solving are modeled
on the gestalt concept of microgenesis. Perceptual microgenesis is a
fact of perception, the qualitative unfolding of percepts over very
short durations. It shows a structure of increasing differentiation
that is useful for discussing creative thought. Such problem solving
takes a similar route, but is regulated by the motivational desire to
decrease tension. There can be forward moves and backward steps,
and these can be modeled as dynamic cases of satiation and then
reorganization. Case studies from Arnheim’s discussions of writing
poetry and Picasso’s working habits are presented.
In Chapter 11, “Individual Artistic Development,” Arn-
heim’s theories of childrens’ art are framed within a life-span per-
spective of continuing artistic differentiation. In general, the indi-
vidual passes from a perception and depiction of generalities to a
conquest of reality, and finally a contemplation of that reality. Mo-
tivation drives the artist over their life-span with a desire to make
their actions correspond with their self-image. The earliest stages of
artistic production are outlined from Arnheim’s writings on child
art, and supplemented with the writings of Claire Golomb and Lucia
Pizzo Russo. The importance of the representational concept is
reiterated, and the necessity of a theory of differentiation that does
not commit the intellectualist error in regarding circles as concrete
objects but instead generic ‘things.’ The development of objects
and compositions are outlined, with fruitful analogies to the com-
positional principles of The Power of the Center.
Individual differences and pathologies are next considered
in Chapter 12, “The Dynamics of Differential and Psychopatho-
logical Artistic Form.” The individual is modeled as a dynamic
system with certain systemic tendencies that can determine individ-
ual differences. The consequences for the creative individual are
6

7
sketched, as supplemented with the consonant theories of Michael
Wallach and Howard Gruber. Next, organic psychoses are discussed
with the particular case of autism, and functional psychoses with the
case of schizophrenia. Although unrelated etiologically, the artistic
productions of the two pathologies represent an interesting hyper-
trophy of the basic formative principles of art making. Finally,
neurosis and art therapy is treated. The mandate to make the reality
of the patient’s artwork conform to the shared reality of sanity is
stressed.
Chapter 13, “The Dynamics of Cultural-Perceptual Form,”
is a discussion of the effects of culture on the creation and recep-
tion of art, without suggesting that cultures develop according to
any particular pattern. While cultures select for diverging percep-
tual skills according to ecology and institutional practices, the im-
portant thing is that skills and the kinds of art that result from them,
are lawfully determined. This point has been underutilized in dis-
cussions of art and culture.
In the final chapter (14), “Objective Percepts, Objective
Values,” the preceding three chapters are used as temporal frames
of reference to analogize to the gestalt nature of the spatial frame
of reference utilized in the second part of the book. In this way,
potentially relativizing aspects of perceptual learning – set, emotion,
mental development, personality and the like – are isolable and
help one discover the ‘objective’ percept. This then leads one to
objective values, which are the reason art has a part of our lives in
the first place. Arnheim argues that values are always instrumental
and require that we specify the context in which we find value.
Value can be discovered, and confirms the role of psychology as an
underlaborer in the understanding of cognitive life.

8
CHAPTER 1
ARNHEIM AND DIFFERENT APPROACHES TO THE PSY-
CHOLOGY OF ART
The ‘psychology of art’ attempts to apply psychological principles
to art. Psychology is – or attempts to be – a ‘science,’ and so the
psychology of art is the application of scientific ‘principles’ of
psychology to art. Aesthetics and all ‘critical’ approaches to art are
(typically) based on principles of logic and argumentation, and in
this sense, the psychology of art can never be the only approach to
art. To be sure, many have turned to psychology as a way of get-
ting away from the subjectivity of ‘mere’ aesthetics, but based
upon a division of labor, there are certain matters upon which psy-
chology can never make the final conclusion. Perhaps the best we
can hope is the injection of psychological debate into aesthetics;
but psychology will not envelop aesthetics, as Comte hoped sociol-
ogy would envelop history.
Terms like ‘science,’ and ‘psychology,’ are by no means
self-evident principles. Even a synthetic definition has the advan-
tage of providing a standard of consistency within a particular
work, regardless of the fact of its truth. Here, ‘science’ is defined as
the systematic, experimental search for universal principles applica-
ble to the organic and inorganic world. ‘Psychology’ is the scien-
tific search for universal principles applicable to human behavior
(Mandelbaum, 1984, pp. 158-170).
In our definition, psychology differs from sociology in the
fact that the data studied by sociologists (when they are truly so-
ciological) are based upon how societies differ from one another;
psychology, on the other hand, possesses data applicable to every-
one (Mandelbaum, 1984, pp. 171-183). To take the example of
socialization of class, class itself is sociological, whereas the univer-
sal way in which individuals adapt to roles and norms of society is
psychological. The psychology of art leaves to the sociology of art
analogous problems. How an artist's guild arises in a society is a
sociological question, how an artist deals with individuality and
conformity to guild structure might in large part be psychological.
The psychology of art is as old as psychology itself. Gustav
Theodor Fechner, one of the founders of the psychophysical
method, published his Vorschule der Ästhetik in the mid-nineteenth
century in which he reported studies on preferences for forms. As
Arnheim (1980) has written, at least three of the earliest, influential
texts on aesthetics were by psychologists. In addition to Fechner's
Vorschule, these included Theodor Lipps's Ästhetik (1901) and
Herbert Langfeld's The Aesthetic Attitude (1920). As we shall see

shortly, some of these original approaches are still exerting influ-
ence in the present day psychology of art.
It is my contention that Arnheim's perceptual psychology
of art serves a fundamental need in studies of the psychology of
art. We need such a psychology of art because until a rigorous psy-
chology exists for the simple level of perceiving, it will be difficult
to apply more developed principles that rely on more fundamental
levels of psychological functioning, like the personality and proc-
esses of motivation. Arnheim's many criticisms of psychoanalytic
interpretation can be seen in this way. That is, he reacts to it the way
he does not so much for its inferior explanation for an identical
phenomenon, but rather because of its scientific positioning.
Psychoanalysis deals with the working of the unconscious in
the artistic works of artists. Psychoanalytic theories have come a
long way from so-called ‘vulgar Freudianism’ in which pointed
objects are interpreted as phalluses, and concavities as female geni-
talia. These theories, in which everything was straight-jacketed into
the schematism of the Oedipal situation, were criticized by Arnheim
many years ago.
One of the most fruitful avenues to pursue lately is object
psychology. Melanie Klein shifted attention from the Oedipal
situation to the pre-Oedipal world of mother-child relations. Adrian
Stokes has been the most significant to apply her ideas. Also im-
portant in object-theory is Winnicott's idea of transitional objects.
Peter Fuller, Richard Kuhns, and J. Randolph have suggested uses
of such Winnicottian ideas and Arnheim (1992, p. 13) has even
admitted that works of art can be treated as such ‘transitional ob-
jects.’ Finally, there is the theory of Jacques Lacan and his follow-
ers.
What Arnheim would object to is a somewhat naive episte-
mological idea common to many psychoanalytic treatments that
sees unconscious drives as working blindly without feedback from
the environment. Furthermore, Arnheim has been disappointed with
the inability of most psychoanalysis to respond to the greatness of
works of art. Arnheim points out that Freud saw his theories most
applicable to popular art, as for instance, pulp novels, and the like.
In "Creative Writers and Day-Dreaming” (1908/1985), Freud saw
the dream as a case of wish-fulfillment, and compared works of
lesser art to similar kinds of wish-fulfillment.
Because Gestalt psychology attempts to explain many as-
pects of meaningful perception at an elementary level, it promises
to serve as an important fundamental level of analysis. If, as Arn-
heim argued many years ago, a round form represents for a child
‘thingness,’ it cannot be interpreted as representing some other
idea that psychologists might like to foist upon the child. Without a
psychology sensitive to such factors, the speculative psychologist
9

will operate at a level of generality that is too great to exhaust cer-
tain formal characteristics of the work that are necessary to dispense
with before further speculation is possible.
Therefore, regardless of whether we follow Arnheim and
use Gestalt psychology, perceptualism is extremely important for
the psychology of art. I shall review the contemporary situation in
perception, but first let us ask how much of a right we have seeking
out some kind of gestalt approach today.
Cognitive Nativism vs. Cognitive Inferentialism
If Arnheim's theory can be classed as a perceptualist approach to
the psychology of art, there are distinct advantages it has over other
approaches. For one, Gestalt theory's blending of realism and the
formative power of the human mind strikes a compromise between
two extreme perceptual views that I am calling Cognitive Nativism
and Cognitive Inferentialism. Some of the theories I will be dis-
cussing are not often classed together and our failure to think of
them together has blinded us to the real advantages of gestalt theo-
rizing. But when the alternatives are cast in the way that I shall cast
them, then the gestalt alternative emerges almost as a necessary
antidote.
Cognitive Inferentialism and Cognitive Nativism are differ-
ent ways to name Helmholtzian inferentialism and Gibsonian direct
theory, the two alternatives in perceptual theorizing. But the terms
are chosen to widen the net. Take Gibson's theory for example. He
has a particular theory of perception that emphasizes the direct
perception of reality through the ‘pick-up’ of invariants in the
environment. But this approach is also compatible with sensory-
physiological theories of perception based on lateral inhibition
(Jameson & Hurvich, 1975; Jameson, 1989; Ratliff, 1992).
Chomskyan innatism and also the ethology of Konrad Lorenz are
compatible. All three theorists stress native abilities and while the
latter two believe in some kind of rationalism, the important point is
that each stresses unlearned abilities and the sufficiency of stimula-
tion for perception and thought.
Likewise, Helmholtz's theory is based on ‘unconscious
in-ference’ which in Helmholtz's own guise was based partly
on judgment and past experience but has been rehabilitated by
Irvin Rock, Julian Hochberg and Richard Gregory to work on the
computation of incoming stimuli. But the approach is also
similar to various theories emerging from the Cognitive Sciences.
One of the virtues of this classification, in fact, is to stress that the
promise of Gestalt psychology was not necessarily fulfilled with
the cognitive revolution. ‘Cognitive’ can mean many things. But
regardless, Inferentialism argues for the lack of sufficiency of
perception to guide perception and thought, hence cognitive
operations.
10

There have been many criticisms pointed at both schools,
loosely conceived. The Gibsonian school is anti-naturalistic in not
wishing to discuss physiology and neural functioning. It is simply
bracketed out, until an ‘ecological optics’ can be developed (Pren-
tice, 1951; Arnheim, 1952; c.f., Epstein & Park, 1964; Epstein,
1977; Proffitt, 1999). Problems remain when perceptual pheno-
mena like lightness and co-planarity are ‘coupled’ and are not
picked up directly by stimulus relationships (Epstein, 1982). An
interesting fact, unforeseen by Gibson, has recently come to light.
With all his emphasis upon higher order variables, Gibson never
stopped to consider that higher order variables, themselves, could
be ambiguous in the sense that first order variables are. But after
examining the projective geometry of certain stimulus arrays, Cut-
ting (1986) has found that there are often multiple higher order
variables, and that the visual system still has to choose between
them.
But there are further problems, more closely aligned to art
itself. If for Gibson sensation is all there is, we wonder then what
space is left for imagination and the free play of thought. In fact
Gibson has had a notoriously difficult time discussing anything
other than representational art (Arnheim, 1979). Followers of Gib-
son like Hal Sedgwick and Sheena Rogers (1996) have used the
horizon-ratio relation to investigate pictorial depth in pictures to
great effect. However, when we are not seeing a naturalistic rendi-
tion of a scene the theory has little to say. The simplicity solution in
perceiving allows the human mind to see new relationships in what
are admittedly non-adaptive stimuli.
For the Inferentialists, the amount of processing is just too
immense. And there is always the problem of the homunculus;
who is directing the cognitive operations? If perception is like
solving a problem, who determines that it is a problem (Kanizsa,
1985)? On the other hand Inferentialism can discuss most aspects
of art, but that is precisely the problem. New cognitive operations
and taking-into-account can always be hypothesized and modeled.
Gestalt psychology defines perception as a problem of per-
ceptual organization. Depending on prevailing conditions, the
stimulus is organized into the simplest percept (according to known
laws of physics). This makes perception neither cognitive nor
homuncular, nor ungrounded in physical principles. While the con-
cept of organization has always been accused of being vague, it has
advantages over both Gibsonian and Helmholtzian approaches, and
especially in the areas they cannot explain. Both simple images
under reduced conditions of stimulation and rich images with ocu-
lomotor cues absent can be handled with relational information
organized according to a simplicity principle.
11

At the most elementary level, say, with two points of light
moving perpendicularly to one another, the theory simply predicts
that the two will be seen to be moving toward and away from each
other. As complexity increases, the number of higher order vari-
ables likewise increase (and we recall that Gibson got this idea from
Koffka, his senior colleague at Smith College for several years).
Here oculomotor cues become irrelevant because there is so much
relativistic information available. In this way, impoverished labora-
tory situations as well as everyday perception can be explained.
Here what is usually described as a flaw in Gestalt theory is
actually its strength. What I call below the "notorious brain model"
(ch. 4) of Gestalt psychology usually discounts it from serious con-
sideration. But reflection on molar physical processes capable of
reflecting isomorphically units of perception and behavior was
intended to bridge precisely the gap between the realism-without-
mechanisms of the Gibsonian school and the infinite-regress-of-
processing of Inferentialism.
Moving on to the arts, the gestalt approach has other dis-
tinct advantages. The sensory-physiological school works with
mechanisms that are simply too primitive to be the carriers of the
depth of content of works of art. Helmholtzian theory does not
have this problem because associations and knowledge can be called
into play in interpretation. But this goes to the point of triviality,
because both normal perceiving and understanding a great work of
art are no different. Gibsonian theory has troubles in that being so
literalistic, it is only comfortable with naturalistic art. Simple per-
cepts found in non-representational painting are analogous to those
experimental set-ups hated by ecological theory. But Gestalt psy-
chology can not only explain their perception, but also point to the
ways in which meaning can arise in them.
Alive or Dead? The Gestalt Approach to Perception
All would agree that Arnheim practices a ‘Gestalt Approach to
Perception,’ but beyond his own ideas can such a ‘Gestalt ap-
proach’ be said to exist? I believe it does and that it is necessary to
expose it so that Arnheim's writings make sense and find their de-
fense. In fact, whenever Gestalt theory has defended itself against
attacks (Köhler, 1959; Luchins, 1953; Rock, 1960; Henle, 1990),
this has indicated that at least some scientists considered themselves
‘Gestaltist.’
As a matter of fact, the survival of the Gestalt movement has
been in question since the untimely deaths of Koffka and
Wertheimer in the 'forties. Around 1960, Wolfgang Metzger re-
sponded to the charge that the Gestalt approach to perception was
‘überholt.’ Since the retirements of Metzger's second generation of
Gestaltists, the rubric has progressively weakened.
12

13
I would like to try to demonstrate that the Gestalt approach
has historically been a unified approach to perception. This can be
accomplished by reviewing a little history. First of all, no one ques-
tions the unity of the school previous to 1945, that is, previous to
the end of World War Two. But as already mentioned, during the
war Koffka (d. 1941) and Wertheimer (d. 1943) passed away. And
even though Kurt Lewin (d. 1947) is most clearly associated with
social psychology, his passing soon after the war served to weaken
the unity of the school.
Köhler, himself, continued publishing into the 'fifties and
'sixties. But by this time his students carried on the work of the
school. Köhler's students, like Arnheim himself, included ‘Gestalt’
in the title of scientific papers. Arnheim, Hans Wallach – repre-
senting those trained in Berlin – as well as Solomon Asch, Mary
Henle – all contributed to the Sourcebook of Gestalt Psychology
(Henle, 1961).
In America, Wallach was only joined by William Prentice
(c.f. 1951, 1956, 1959) and Nicholas Pastore (c.f. 1956, 1960) as
true perceptionists associated with the gestalt school, and this would
seem to weaken the idea of the continuity of the school. Character-
istically, Prentice eventually left experimental psychology for uni-
versity administration, but not before contributing the important,
"The systematic psychology of Wolfgang Köhler" (1959). But this
is to ignore Europe – particularly Germany and Italy – where Ge-
stalt psychology flourished during this time, and Arnheim was very
close to his colleagues in Europe.
In Germany, Wolfgang Metzger, Edwin Rausch, Wilhelm
Witte, among others, remained through and after the Second World
War. Already in 1936, one year after the publication of Koffka's
The Principles of Gestalt Psychology, Metzger published the first
edition of Gesetze des Sehens [The Laws of Vision] (1975), which
would see two more editions (1953 and 1975). Furthermore, in
1941, he published his general text, Psychologie, which also went
through several editions (Metzger, 1941). Around the same time,
Edwin Rausch published one of the most definitive works on geo-
metrical illusions, Struktur und Metrik figural-optischer Wahrneh-
mung (1952).
In Italy, the Meinongian Vittorio Benussi had taken a post
after the First World War, and introduced experimental psychology
into the country. His student Cesare Musatti took over the chair in
Padua, which Benussi had occupied before his suicide. There, Mu-
satti steered his ideas in the direction of the Berlin gestalt theory,
and through his students and near contemporaries – Fabio Metelli
and Gaetano Kanizsa – formed a triumvirate of researchers.
By the 'fifties, Musatti in Milano, Metelli in Padua and
Kanizsa in Trieste formed the nucleus of Italian Gestalt psychol-

ogy. Needless to say, the German and Italian psychologists were
extremely close and it is said that Metzger – somewhat reminiscent
of Koffka coming up to Berlin from Giessen (c.f., Heider, 1983) –
would visit from Germany and tour the laboratories of Metelli and
Kanazsa with his friend, Musatti. While Musatti left experimental
psychology in the 'fifties for more speculative work, Metelli and
Kanizsa remained to represent Gestalt psychology (Verstegen,
2000a).
In the late 'fifties, a new generation of students began ap-
pearing – the theoretical ‘grandchildren’ of Wertheimer, Koffka
and Köhler. In America this was primarily at the New School for
Social Research, where Arnheim taught along with Solomon Asch,
Mary Henle and Hans Wallach. In Germany, the psychological
Institute of the University of Münster was the center of Gestalt psy-
chology, where both Metzger and Witte taught. And as mentioned
before, Metelli was in Padua and Kanizsa in Trieste.
Notable American students were Dorothy Dinnerstein and
Irvin Rock, who were the first to earn their doctorates at the New
School after the Second World War. Other students were John
Ceraso, Sheldon Ebenholtz, and William Epstein. In rare cases, stu-
dents like Ceraso – who was a research assistant to Solomon Asch at
Swarthmore – had direct contact with Köhler. In the mid-sixties,
doctorates were granted by the Institute for Cognitive Studies, Rut-
gers, where Asch, Dinnerstein, Howard Gruber, and Rock taught.
It is characteristic that probably none of the students of the
so-called ‘New York school’ would call themselves ‘Gestaltists,’ in
sharp distinction from their European counterparts. Ceraso might
be an exception, and Alan Gilchrist, a product of Rutgers, recently
entitled a paper "Developments in the gestalt theory of lightness
perception" (1990).
In Münster, the most important students of Metzger's were
Suitbert Ertel, Lothar Spillman and Michael Stadler. In Italy, Paolo
Bozzi, Walter Gerbino, Osvaldo da Pos, Marco Sambin, Giovanni
Vicario and Mario Zanforlin, and all began to publish their ex-
periments. It might be said that the Europeans watched with great
interest the emergence of the American cognitive revolution. Much
of it must have seemed superfluous to them, who were possessed of
a greater sense of the history of psychology (Verstegen, 2000a).
The Americans reaction can be attributed to the
unsuccessful cultural transmission of the gestalt school to America.
Gestalt psy-chology always seemed mysterious, and unscientific to
them. Even older psychologists like Fred Attneave and Wendell
Garner shared this view (c.f., Henle, 1990). Perhaps if they
had the intimate knowledge of the earlier and subsequent
German literature that Carroll Pratt or Harry Helson had, they
would have had different opinions.
14

Be this as it may, none of them accepted the gestalt name.
Some, such as Irvin Rock, seem to have devoted their entire profes-
sional careers to coming to terms with their New York training –
Rock, most recently, reverting back to a pure Helmholtzian posi-
tion. But one finds that if one compares the work of some of these
non-gestaltists, their work is indistinguishable from Europeans of
the same age.
William Epstein (1977, 1982, 1988, 1995; Epstein and Park,
1964), for example, would not call himself a gestalt psychologist,
but his thoughtful reviews of theoretical issues in perceptual psy-
chology since about 1960 follow the development of European
gestalt theory itself. His consistent positioning between old-style
inferential theory and direct realist, ecological theory is exactly
paralleled in gestalt theory.
Fig. 1, Wertheimer’s laws as instances of grouping by similarity,
after Arnheim (1974)
Gestalt psychology has experienced a revival, sometimes
under the adoption of the so-called ‘Cognitive Revolution.’ How-
ever, one still finds misunderstandings and a condescending opin-
ion by those emerging from the computer and cognitive disci-
plines. Arnheim has had no small part in trying to correct matters,
emerging as a major interpreter of the gestalt tradition (Arnheim,
1986b, 1992, pp. 200-214; Perkins, 1986). A rapprochement is
15

16
being accomplished by scientists who are coming to appreciate
their historical ties to Gestalt psychology in their schooling as well
as by other scientists, mostly American, who are trying to under-
stand the true meaning of the gestalt tradition. Examples are Mi-
chael Kubovy (1986) and Stephen Palmer (Robertson, 1987).
Relational Determination and Simplicity
Gestalt psychology won its reputation with shape perception, and
every introductory textbook of psychology describes Wertheimer's
famous ‘laws of visual organization’ (Wertheimer, 1923/1939).
Wertheimer enunciated a number of laws – such as the law of shape,
color, proximity and orientation – although as he said in his text he
was only outlining general principles that were in no sense absolute.
Arnheim (1974) reinterpreted these as instances of a common prin-
ciple of similarity (Fig. 1).
More important to Wertheimer was the concept of präg-
nanz, or the clearcutness or pithiness of visual forms. In other
words, given a particular stimulus configuration, we tend to see the
most clearcut, pithy, organization possible. Individual laws are only
subordinate to this general principle. Köhler hypothesized that the
brain was directed by energy minima and this has come to be
known as the ‘simplicity principle.’
The principle of prägnanz has been accused of being vague
and mysterious. Arnheim tried to help matters. But Erich Gold-
meier (1982) has revitalized the theory so well that any criticism
should take cognizance of his theory, and only then can a criticism
be properly leveled. Goldmeier takes advantage of the translation
of singularity, which means that a prägnant form is singular in the
sense that it is the form to which other forms are similar, and not
vice versa. The mysterious tendency toward prägnanz is nothing
more than the tendency toward stability.
The tendency toward stability in the phenomenal field is
also a product of a tendency toward stability in brain processes.
The simplicity principle has been subjected to various attempts at a
metric (Hochberg, Attneave, Leeuwenberg) and has been accused
of being vague because it is difficult to decide on an area where this
simplicity is applied. There is a strong ad hoc quality to the sim-
plicity principle but it has also been fruitfully defended (Hatfield &
Epstein, 1985).
The substance of the gestalt theory of figural perception
can be summed up in the concept of relational determination based
on the field metaphor (Koffka, 1935; Witte, 1966; Metzger, 1975,
chapter 18; Rock, 1990). Gestalt psychology has been led by the
field metaphor, and the determination of perceptual effects by the
state of the rest of the field. Thus the way in which a line is per-
ceived is based upon the influence of the surrounding stimuli.

17
Gestalt psychologists have gone on to explore all of the tra-
ditional problems of figural perception and proposed and experi-
mentally tested their explanation based upon relational determina-
tion. As Rock (1990) has pointed out, some phenomena are more
easily treated with the concept than others, but on the whole it is a
fruitful concept for their explanation. It continues to be important,
for example, in the explanation of figural illusions (Erikson, 1970)
and anomalous figures (Sambin and Pinna, 1987).
Recently, dynamic principles have become the centerpiece
of the theory of perception of Steven Lehar (1999), who bases the
emergent form of an anomalous contour of a Kanizsa triangle, for
example, on the action of vectors (Fig. 2). These various models
are all based on vectors in a field and with this in mind it can be
seen that Arnheim has basically provided his own interpretation of
Wertheimer in his The Power of the Center (1982/1988). This will
be outlined in the next chapter.
Fig. 2. The creation of an anomalous contour (from a Kanizsa
triangle), according to Lehar (1999)
Toward the Explanation of the Meaning of Art
When we think of Gestalt psychology and art, we immediately think
of Rudolf Arnheim, who has already been mentioned. But the basis
of a gestalt approach to art goes back even before him. All of the
founders of Gestalt psychology, which included Arnheim's teachers,
had theories of the psychological foundations of the arts (Koffka,
1940). Among these we must mention Johannes von Allesch, Erich
von Hornbostel, Max Wertheimer. There was also a notable school
of art history, which focused on problems of visual structure and
color relationships (Kurt Badt, Otto Pächt, Hans Sedlmayr, Ernst
Strauss) (Verstegen, 2004).
Along with Arnheim, the most important commentators of
his generation were Victor Zuckerkandl and Carrol Pratt (Arnheim,
1975). Arnheim's remarkable longevity and the distrust of
‘schools’ in the scientific temper of America has discouraged any
followers of Arnheim in America (Verstegen, 1996). Notable, how-
ever, is Claire Golomb at the University of Massachusetts, who

18
works on children's art (c.f. ch. 11). In Europe, however, one must
take note of the writings of Alberto Argenton (1996), Augusto
Garau (1984/1993), Manfredo Massironi (2002) and Lucia Pizzo
Russo (1988, 2004) in Italy and Max Kobbert (1986) in Germany
(c.f., Bonaiuto, Bartoli and Giannini, 1994). The semiotic work of
Fernande Saint-Martin (1990) is indebted to Arnheim, and utilizes
a dynamic methodology very amenable to Arnheim. I shall draw
frequently on these works in the following, which point to a vigor-
ous and lively research tradition that is almost unknown in Amer-
ica.

19
CHAPTER 2
EXPRESSION, SYMBOLISM AND VISUAL THINKING
With the general outline of Gestalt psychology made one can move
to the realm closer to art proper. Indeed, some critics have charged
that there is nothing wrong with Gestalt psychology but rather the
way it is presented by Arnheim that is at fault. Much of the benefit
of the gestalt approach is its ability to deal with problems of expres-
sion.
Many use the arts to enlighten aspects of object perception
but the real domain of the psychology of art, the part that shows
how a psychological model can do justice to the profundity of the
human mind as well as its artifacts, is expression. But there is still a
gap between the symbolic and expressive life of people and their
creations.
How is emotional life related to the experience of objects?
Is it independent or is it related; is it dominating or parasitic? To
begin we can insist on some sub-stratum of meaning that operates
cross-modally. One of the most utilized methods of attacking this
expressive life is with the tool of the semantic differential. Devel-
oped by Charles Osgood (Osgood, Suci & Tannenbaum, 1957;
Osgood, May and Miron, 1975) and put to good effect in his neo-
Behaviorist psycholinguistics, it is a useful tool. In the work of Dan
Berlyne and his school it has been used extensively for under-
standing expression. And the gestalt psychologist Suitbert Ertel
actually worked with Osgood, producing impressive variations on
the semantic differential technique (Ertel, 1964, 1969).
Although it is a powerful statistical tool, some important
criticisms have been leveled against the technique by gestalt-
oriented psychologists based on its irrationalist bias. Work by Dean
Peabody (Peabody & Goldberg, 1989) has pointed to its inability
to separate descriptive traits from emotive traits. Since descriptive
traits are cognitive, it fails to show how emotion derives from cog-
nition. Ertel’s and Peabody’s insights ought to be brought together
to form a more adequate statistical tool that can uncover real un-
derlying structural patterns, rather than statistical artifacts.
More promising is some sort of multi-dimensional model
of expressive lexical terms based on psychological poles of real
phenomenological meaning. Arnheim (1962/1966) distinguished
three levels in the perception of affects. First, he identifies a crucial
cognitive stage of the identification of objects. This could corre-
spond to the affect as it is experienced. Then, there is the expressive
and motivational component, which are identical. This could corre-
spond to the perceptual expression which is available to other per-
ceivers. Finally, there is the ‘emotional,’ the level of tension of the

act as it is perceived by the person themself. The positions of the
psychologists of emotion, Joseph De Rivera (De Rivera, 1977; De
Rivera & Grinkis, 1986) and Bernard Weiner (1986) are in large
agreement with Arnheim. From the computational side of psycho-
logical research, there are interesting corroborating developments
(Thagard & Nerb, 2002).
They, like all gestalt thinkers, take it for granted that an af-
fect is fundamentally a relationship between two things (c.f., Asch,
1952; Heider, 1958). It is no accident that Arnheim's model in The
Power of the Center treats centers and their eccentric centers almost
on an individual model. Just as individual affects arise in relation-
ships between people (‘anger,’ ‘disgust’), so too artistic expression
arises in the vectoral relations between centers. And just as affect is
the perceived attributional relations between actors, expression is
the outcome of perceptual organization of perceived units.
In his essay, "Emotion and feeling in psychology and art"
(1966), Arnheim is always careful (as I was above) to use the ra-
tionalist term affect that is found in Descartes and Spinoza. ‘Emo-
tion’ was left as a subsidiary level of excitement pertaining to the
rational basis of affect. Arnheim's reforms have not been followed
in general psychological theory where the term ‘emotion’ retains
its popularity, however, they are particularly apropos in artistic
studies because when we view art, we do not experience genuine
emotions.
This terminological precision points to Arnheim's dissatis-
faction with a term like ‘aesthetic enjoyment’ or ‘aesthetic emo-
tion.’ These trace aesthetic experience back to hedonistic inter-
pretations of art. But if art is a much more serious matter, concern-
ing cognitive material, then the semantic message rather than a
mere sensation is at issue.
Expression
It is a feature of Gestalt thinking to grant the expressive value of
percepts. The world issues a ‘requiredness,’ and expressive qualities
communicate this. Thus Gestaltists have tried to elevate the mistak-
enly named ‘tertiary qualities’ (Pratt, 1962; Bozzi, 1990). Gibson’s
idea of affordance has tried to retain something of this meaning but
characteristically the meanings it has captured are largely literal.
Gibsonian affordance tells us when we see a woman but they cannot
make us feel, as Koffka said, ‘Love me.’
There has been little effort to grant the inherent expressive-
ness of artistic form. The rebirth of empathy theory (e.g., Crozier
& Greenhalgh, 1992) simply shows how scientists are only willing
to look elsewhere to grant expression. Recently, Arnheim (1988b)
invoked the research of Heinz Werner’s ‘sensory-tonic’ theory of
perception (e.g., Werner & Wapner, 1954) to bolster his own, but
20

21
once again this location of expression in the tonus of the body and
not in percepts themselves is unduly weakening of his true position.
In the same way that Arnheim is suspicious of emotion and
feeling, he is suspicious of an uncritical invocation of physiog-
nomic perception as the basis of the arts. This marks a difference
between Arnheim and theorists he and Gestaltists are often grouped
with, like Heinz Werner. Physiognomic perception, while certainly
quite real, implies regression, syncretism and instead Arnheim
nominates metaphor-like processes that recognize the isomorphic
structural qualities of diverse sensory situations (ch. 9).
Arnheim does not derive metaphor from physiognomic
perception but rather vice versa, physiognomic perception from
metaphor (Glicksohn & Yafe, 1998). In this he shares similarities
with recent work by Michael Wallach, Nathan Kogan and George
Lakoff. Arnheim would agree with Wallach and Kogan's (1965)
statement that "to respond to the physiognomic properties of things
and events involves an act of metaphor, an act of simile, or an act of
signification" (p. 144).
The gestalt position is slightly different, however, from the
cognitive position of Lakoff. Lakoff founds metaphoric analogies
on mediated associations that he calls ‘grounding, but which are
relatively physicalistic or behavioristic. They refer to the coinci-
dence of features rather than structural similarity. This has conse-
quences for the way a metaphor is apprehended. For Lakoff, when
a metaphor is invoked, you are left with the stock response (e.g.,
‘life is a journey’) and you are done (Tsur, 1999).
Cognitive or structural models like that of Clark and Clark
(1977) are better. In such a system unmarked features are valued as
positive, and marked as negative. The unmarked features at the
ends of the analogous scale correspond to each other: thus, fast
vibrations are perceived as 'high' tones, slow vibrations as 'low'
tones. Greater height and more frequent vibrations are the un-
marked (more salient) extremes of their respective scales, and thus
they are matched.
Gestalt psychologists have traditionally emphasized this
doctrine of the ‘Unity of the Senses’ (Hornbostel, 1939; c.f.,
Marks, 1978). Thus there are structural affinities between the dif-
ferent senses and when they are compared we can metaphorically
note these affinities. This is just another expression of the gestalt
idea of isomorphism, except that it does not relate to the perceptual
and electrochemical levels but between perceptual realms.
Not unlike physiognomic perception, synaesthesia has also
been equated with metaphor (Marks, 1978). As has been pointed
out by Kennedy, this disrupts the traditional asymmetrical relation-
ship between vehicle and topic of metaphors (Kennedy, 1990). It is
best to regard synaesthesia as a genuine phenomenon of human

22
perceiving that is based once again on real structural affinities be-
tween the senses; however, it must proceed metaphorically.
Abstraction, Symbolism and Visual Thinking
When a metaphor is active, it plays with levels of abstraction in or-
der to lay bare a deeper structural affinity between two perceptual
images. Thus the structural features are the raw given in the envi-
ronment that allows for symbolization. This is so because the sym-
bol, according to Arnheim (1969), "portrays things which are at a
higher level of abstractness than is the symbol itself" (p. 138).
Such symbolism is common in all the arts but differs in its
use due to the nature of the individual media. Language, for exam-
ple, seems to be somewhere between music and vision in terms of
abstraction. It is characteristic of Arnheim’s thinking that the no-
tion of a symbol is completely perceptual and he dispenses com-
pletely with any kind of logical semiotic classification. Anticipating
Peirce’s discussions of icons, Arnheim responds that images are
pictures "to the extent to which they portray things located at a
lower level of abstractness than they are themselves" (p. 137). This
is his solution to long-standing semiotic difficulties relating to sign,
icon and index. It is not based on arbitrary criteria but rather levels
of perceptual abstraction; symbols are of a higher level of abstrac-
tion, and pictures a lower level of abstraction, than the thought they
represent.
There are two competing aspects of Arnheim’s thinking on
‘the intelligence of the senses’ here that have never really been
acknowledged. There is on the one hand the ability of the senses to
contain universal or abstract information. And there is on the other
hand, the manipulation of images for productive thinking. We
might relate the two by saying that individual percepts already
contain abstract content, just as a work of art can be called the ab-
stracted solution to an artistic problem. It is, however, the manipu-
lation of symbols within the work of art that represents the prob-
lem-solving aspect of creation and the means to the solution of the
final work. I will briefly touch on these two distinct issues.
In a classic paper, "Perceptual Abstraction and Art" (1966),
Arnheim considered the ‘intelligence’ of the senses by pointing to
a fact now taken for granted in Cognitive Psychology, the abstract-
ing nature of perceiving. The gestalt theory of prägnanz takes care
of abstraction because in providing stable nodes with which the
mind can economically organize stimuli, it creates the ability to
hold many examples under one rubric. As Arnheim says "all per-
ception is the perceiving of qualities, and since all qualities are ge-
neric, perception always refers to generic qualities."
When Arnheim wrote his article his was a lone voice. Now
there are many more excellent approaches, dealing both with per-

ception (Stephen Palmer) and cognition (Eleanor Rosch). Prägnant
or ‘singular’ perceptual forms or cognitive objects are exemplary
and dictate the asymmetrical relationship between these objects
and those that are similar. However, the old linear model of
monotonic dependence between variation of stimuli and
phenomenal qualities still persists in some cognitive psychology
(Zimmer, 1986). The difficulty of quantifying prägnanz once
again frustrates cognitive psychologists but it is a necessary
concession to the facts.
By its very nature perception contains abstraction but it also
possesses another form of intelligence in the spatial opposition of
agents to embody patterns of thought. Rather than relying on ma-
nipulation of codes, Arnheim has been a strong proponent that
productive thinking relies on the manipulation of spatial variables.
Thus in the case of a logical problem, spatial modeling provides the
synthetic judgment of the necessity of the solution. Arnheim gives
the example of the logical syllogism: If A is taller than B, and B is
taller than C, is A taller than C? This opacity of the words hinders a
successful solution of the problem, until it is thought of in spatial
terms. Arnheim’s ideas have been confirmed by other researchers
(Huttenlocher, 1968).
Today, there are several different theories that go on to use
some sort of spatial analog of a problem situation and look to its
manipulation in movement for the development of analogies, cate-
gorization, and other productive outcomes (Croft & Thagard,
2002). Similarly, the use of diagrammatic reasoning is another big
research interest that may amplify Arnheim’s position. Knauff and
Johnson-Laird (2000) found, for example, that easy visualization
was not so responsible for reasoning as easy spatialization. The
most promising model of problem solving today is the theory of
mental models (Johnson-Laird, 1998). The creation and manipula-
tion of mental models allows for productive thinking. The question
arises as to whether or not it is compatible with Arnheim’s emphasis
on visual thinking. Johnson-Laird distances his idea of models
from mental images but it is possible that the two are speaking of
the same thing, for when Arnheim translated his book into German,
he chose the more generic term for ‘visual’ to be Anschaulich.
The dominant paradigm in research on thinking is certainly
information processing. There is some hostility between this ap-
proach and traditional gestalt thinking, because the former in its
classic guise is about the manipulation of elements according to a
preexisting code, rather than being adaptable to novel (‘produc-
tive’) situations (Wertheimer, 1985). Arnheim’s Visual Thinking
(1969) provided much evidence for this case, and was used promi-
nently by Hubert Dreyfus in his influential studies, What Computers
Can’t Do (1972) and What Computers Still Can’t Do (1992). The
challenge of the solution of novel solutions will be with computer
23

24
models for a long time. But lest we believe that Gestaltists are sim-
ply against computers, the important point is that a computer is
needed that supersedes binary operations – regardless of how fast –
in the direction of complex networks. The advent of connectionism
is a start but the most promising model so far is the synergetic
computer developed by Herman Haken and his collaborators.
The senses are intelligent in two ways: by posing objects in
relationships that reveal productive solutions to difficult problems
and by already classifying objects according to abstract principles.
In either case, the relationship of elements that promote under-
standing derived from the cognition of this relationship reveals
expression, which lies in the juxtaposition of unalike elements (sen-
sory objects, mental images).

25
CHAPTER 3
THE SENSES, PERCEPTUAL OBJECTS AND THEIR DY-
NAMICS
Arnheim presumes to offer universal principles for the analysis of
the various arts, but there are obvious differences between various
media (painting, music, literature) and the sensory modalities
through which they are communicated, principally hearing and
vision. Arnheim’s early work on the parallel worlds of film and
radio provide a good paradigm from which to consider these dif-
ferences. His pioneering essay, “A forecast of television” (1957),
addresses this question so directly, I shall cite it at length.
The eye gives information about shape, color, surface
qualities, and the motion of objects in three-dimensional
space by registering the reactions of these objects to light.
The ear reveals little about the objects as such; it only re-
ports on some of their activities, which happen to produce
sound waves. On the whole, the eye takes little interest in
the nature, place, and condition of the light sources that
make the light rays fall upon the retina. The ear is inter-
ested in the source of the sound; it wants the sound waves,
on their way to the ear-drum, to be as little modified as pos-
sible in order to keep the message from the source unal-
tered. Sound is produced by an object but tells us little
about that object’s shape, whereas the eye, in order to ful-
fill its task, must reckon with the fact that a suitable likeness
of a three-dimensional object must be at least two-
dimensional. Any sense organ can register only one stimu-
lus at a time so that the eye in order to produce a two-
dimensional recording has to consist of numerous receptors
that operate one next to the other. The mosaic that results
from this collaboration of the receptors depicts three dimen-
sional space and volume as best it can. The time dimension,
which is available in addition, uses the change in stimula-
tion in each receptor to record motion and action.
A different situation is found in hearing. The sounds that
exist in auditory space at any one time are not recorded
separately but add up to one, more or less complex vibra-
tion, which can be received by a single membrane, such as
the ear-drum. This unitary vibration may be produced by
the simple sound of a tuning fork or the complex noises of a
crowd of excited people or a symphony orchestra. To some
extent the ear succeeds in teasing the complex vibration
apart, but it offers scant information about the locations of

26
the different sound sources. The ear, like the eye, operates
with a battery of receptors, and they, too, are arranged in a
two-dimensional surface. The receptors of the cochlea are
parallel fibers, as different in length and tension as the
strings of a harp, and apparently for a similar purpose. The
‘strings’ of the cochlea seem to be activated by resonance
when vibrations of corresponding frequencies impinge
upon them. This means that the ear uses its receptor field to
distinguish between pitches, whereas the eye uses it to dis-
tinguish between spatial locations (p. 157).
According to Arnheim, then, the eye tells us about things
and their relations, while hearing reports exclusively on what things
do; it is silent about what they are otherwise. “A bird, a clock, a
person exist aurally only as their singing, ticking, speaking, weep-
ing, or coughing; they are characterized only by their adverbial
properties and exist only as their properties endure" (1986, p. 67).
Vision deals in concrete objects, hearing in forces.
Perceptual Objects
These givens about vision and hearing are the starting point for art.
But it is nevertheless possible to speak generally of ‘perceptual
objects,’ and in addition general principles of ‘unity,’ ‘balance’
and finally ‘dynamics.’ The problem manifests itself in
Wertheimer’s laws of grouping, which develop the way units are
formed in mental representations and, in fact, Wertheimer’s princi-
ples were originally intended to cover both the visual and auditory
realms. At the end of his classic study of stroboscopic motion
Wertheimer makes reference to the similar problem of phenomenal
grouping in musical perception (Wertheimer, 1912/1961). The
musicologist Victor Zuckerkandl (1956, p. 136), in fact, goes so far
as to say of Wertheimer’s study that "often we should have to sub-
stitute 'tone of a certain pitch' for 'thing in place' and we should
have a perfectly valid statement concerning heard instead of seen
motion."
There are two issues here, the identity of gestalt laws for
seeing and hearing, and the representation of works of art as a spa-
tial mental image. As for the first problem, any problem of tempo-
ral organization can be described with spatial concepts. Figure and
ground effects can be achieved with both vision and hearing, as can
any number of perceptual illusions. Giovanni Vicario (1982) has
further demonstrated this convincingly in elegant demonstrations
leaving no doubt that temporal organization refers ultimately to
spatial organization. Kubovy and Van Valkenburg (2001) concur
with the joint application of gestalt laws in vision and hearing as
applied generically to ‘auditory and visual objects.’

27
The formality of Wertheimer’s gestalt laws was prominently
contested by the psychologist Geza Révész (1937). Working in
hearing and haptics, Révész questioned the simple extension of
organizational ideas of visual perception to other sensory modali-
ties, thus denying, for instance, that hearing was spatial. Interest-
ingly, Gombrich has cited Révész's criticisms as an indirect criticism
of Arnheim (Sacca, 1980-1). The gestalt psychologist Wolfgang
Metzger (1953) did make a response in which the universality of
gestalt laws to sense modalities other than sight was reaffirmed.
There is some confusion because there is also a tendency
by gestalt theorists to affirm the spatial character of tones. Révész
was especially opposed to this, and the affirmation of a Hörraum
might have seemed to slight hearing in favor of vision (rather than
simply affirming analogous perceptual laws of grouping). However,
there seems to be good evidence for this dimension of auditory
percepts. The gestalt theorist Erich von Hornbostel was the most
prominent to affirm this (Hornbostel, 1926; Zuckerkandl, 1956, ch.
15, Bozzi and Vicario, 1960).
This is still distinct from the second problem of the ten-
dency to form spatial mental images of temporal percepts. Works
of art must have some sort of visual representation in order to be
understood. It must be understood as a complex of pure interac-
tions of forces. This is of course true of static works of art, like
paintings. But is equally true of temporal works of art. The sym-
phony or the novel is perceived as organized wholes when grasped
in their simultaneity. In order to press his point, Arnheim (1974)
asks, “When the dancer leaps across the stage, is it an aspect of our
experience, let alone the most significant experience, that time
passes during that leap? Does she arrive out of the future and jump
through the present to the past? And exactly which which part of
her performance belongs to the present? The most recent, elapsed
second of it, or perhaps a fraction of that second? And if the whole
leap belongs in the present, at which point of the performance does
the past stop?” (p. 373).
Time is then a qualified aspect of the experience of a tem-
poral work of art. What rather happens is that the viewer of a dance
continually relates aspects into a larger whole. “While listening to
music, the hearer weaves relations back and forth and even coordi-
nates phrases as matched pendants, e.g., in the return of the minuet
after the trio, although in the performance they are delivered one
after the other” (1986, p. 71). ‘Matching pendants’ occurs when-
ever two successive parts are similar, and they collapse into a single
symmetrical unit.
There is much evidence that this in fact how artists of tem-
poral works work. Arnheim (1974) cites a famous letter attributed
to Mozart, in which the composer discusses thinking of a work of

music: “[The theme] becomes larger and larger, and I spread it out
more and more widely and clearly, and the thing really gets to be
almost completely in my head, even if it is long, so that thereafter I
survey it in my mind at one glance, like a beautiful picture or hand-
some person. And I hear it in my imagination not in sequence, as it
will have to unfold afterward, but, as it were, right away all together
(wie gleich alles zusammen)” (p. 374, his translation). The musi-
cologist Victor Zuckerkandl (1973, 22) has also said of music, "Can
a gestalt come about anywhere but in space, where it unfolds with
all its parts all at once in simultaneity and where it offers itself to
the observation without receding immediately?" And Arnheim
points out also that Heinrich Schenker's concept of the Urlinie "is
an eminently visualizable notion" (1992, p. 36).
Turning next to unity and balance, it is interesting that the
Gestalt psychologist Wolfgang Metzger (1941), suggested that all
perceptual objects also have a perceptual center. This explains why
we are able at all to see something as an object at all. When a spot is
determined by the visual system at which all relations may be un-
derstood, the stimulus is understood as a thing in its own right. Gio-
vanni Vicario and his coworkers (Beghi, Vicario and Zanforlin,
1982; Davi, 1989; Savardi, 1999) have recently provided a mathe-
matical formalism for determining the perceived center of any con-
figuration. There is also evidence that three-dimensional objects
have perceptual centers of reference. The biological regularity of
plants and humans means that a simple ‘center of moment’ from
which all movements are related is immediately perceivable (Ger-
bino, 1983).
When something purports to be an art object, it has to fulfill
special functions. The artistic theme has to be exemplarily ordered.
Then and only then can it be said to be an example of structural
order. Thus, balance is absolutely required if they are to attain the
status as art objects. “Under conditions of imbalance, the artistic
statement becomes incomprehensible. The ambiguous pattern al-
lows no decision on which of the possible configurations is meant.
We have the sense that the process of creation has been accidentally
frozen somewhere along the way” (1974, p. 20). Works of art re-
quire balance because they require finality to be valid statements on
the human condition.
Some critics have felt uneasy about Arnheim’s elevation of
the necessity of centeredness in artistic composition (Schufreider,
1986). They have insisted, for instance, the center is not even active
in particular compositions. But we have already seen that centered-
ness is a quality of all circumscribed perceptual objects, and (say)
Mondrian paintings do not provide a sufficient counterexample. In
Arnheim’s (1988) words, “as soon as we are faced with a closed
28

29
space…the enclosure mobilizes a field of visual forces that creates a
balancing center and organizes around it” (p. 107).
The basic problem for composition is, therefore, how per-
ceptual centers can interact with a format and find a balancing
center. I have already pointed out (ch. 3) that centricity allows for
unit formation. Whether or not a balancing center is ‘retinally’ or
‘tympanically’ present, they are nevertheless intuitively perceived.
This makes obvious sense with painting, but with the case of a
sculpture, architecture, dance and film, too, the balancing center
must be perceived. Especially in the cases of literature, there is no
physical stimulus to correspond with, and yet it must be perceived
for the work to be comprehended.
The centricity of the composition makes makes the work of
art a unified whole, makes it a readable statement, and provides it
with a measure of finality. When Arnheim speaks of ‘the power of
the center’ he is expressing just this importance. Let me discuss
each of the concepts in turn. The nature of mentally represented
wholes provides a clarifying statement for the arts: “a [work of art]
must be perceived as some kind of visual image if it is to be under-
stood as a structural whole” (Arnheim, 1986, p. 80). There is no
doubt that all works of art rely on time for their understanding, this
suggests that such concepts as time must be more carefully delim-
ited. As I will try to show, Arnheim’s theory cuts through many
critical difficulties about ‘spatial form’ in literature, and time in
paintings, etc.
Perceptual Dynamics
Naturally, the arts become important repositories of thought and
wisdom, visual correlates of complex ideas. However, Arnheim has
gone further to provide a grammar as it were of the way that such
meaning develops. The first assumption of the theory is that all
objects and by extension all artistic objects, visual (the visual arts),
aural (music and speech sounds), or verbal (poetry) are centers of
perceptual energy and attraction. The problem of composition is
therefore how these centers of energy can meaningfully interact.
When Arnheim speaks of such objects, he calls them perceptual
centers. Perceptual centers interact, first of all, with each other, cre-
ating eccentric vectors of action. They, furthermore, interact with
‘frames’ or ‘formats,’ which are really larger perceptual centers
which serve as frames of reference for these relative elements. Any
format can, and often does, become a perceptual center in a larger
frame of reference.
Perceptual objects are centers and are the variably inde-
pendent units with which we are interested. Any perceptual object
“constitutes a dynamic center because it is the locus of forces is-
suing from it and converging toward it” (1988, p. 225). Depend-

30
ing on the scale of magnitude, grouping elements from below we
may arrive at the whole work as a center, or we may subdivide from
above to the elements, which then become centers in their own
right.
Fig. 3. The centric compositional system
To elaborate this with diagrams, let us look at a dynamic
center with vectors emanating from it (Fig. 3). It is a source of en-
ergy, which emanates from its center. Here we treat a vector as, “ a
force sent out like an arrow from a center of energy in a particular
direction. When a system is free to spread its energy in space, it
sends out its vectors evenly all around, like the rays emanating from
a source of light. The resulting symmetrical sunburst pattern is the
prototype of centric composition” (p. 4; see also Arnheim, 1974,
pp. 23-6).
The principal compositional characteristic of centers is
weight. Arnheim (1988) defines perceptual weight as “the dynamic
power inherent in an object by virtue of its conspicuousness, size,
shape, location, etc” (p. 229). We might go further and say that
weight is produced following the rules of organization in any par-
ticular modality. Arnheim (1988) thus gives his first rule for the
compositional effects of weight.
A. Weight increases attraction
Most usually in conjunction with the anisotropy of space, discussed
below, weight is next increased by distance.
B. Distance (I) increases perceptual weight when per-
ception is focused upon the center of attraction.

Arnheim likens this ‘rubber-band effect’ to potential energy in
physics. Weight generated by distance is only increased, however,
when an object is anchored to a base. Otherwise,
[distance] (2) decreases attraction when perception
is anchored in the attracted object.
This means to say that when the object itself is considered as a cen-
ter, it loses its anchoring to the old center. B (1) and B (2) are mu-
tually exclusive ways of looking at a composition and point to the
problem of the dual nature of the dynamics of centers which hinges
on what capacity as center they are perceived to be, and is further
treated below.
The behavior between perceptual centers must take place
within some finite context. This is the larger format. I already men-
tioned that while the limit to which an artist supports or doesn’t
support a format is a matter of choice, the organizing influence of
any ‘closed space’ is not.
A format is nothing more than a large perceptual center
that serves as a perceptual framework for the action occurring under
its perceptual power. This concept is brought out nicely in the
German word Figurfeld. The figure is a field, but the figure is also
a figure, depending on the resolution with which you approach the
relative strength of the various perceptual systems.
The idea of format has found elaborate study in Gestalt
psychology as the frame of reference, or Bezugssystem (Koffka,
1935; Witte, 1966; Metzger, 1975; Rock, 1990). While as an ex-
planatory concept it often remains outside of contemporary ac-
counts of perception, Gestalt psychology holds to its necessity for a
proper account of perception. The frame of reference is only the
most important centric system of the work of art. It is therefore the
‘common component’ to which all others are ‘relative.’ All cen-
tric systems have tonic qualities, but the format is the most impor-
tant for the whole work of art.
The frame of reference is a general psychophysical prob-
lem that extends into the time dimension as well, for instance, to
tempo. It is in the context of any format, therefore, that the compo-
sition must find its balancing center. The format has a special status
because it is ultimately the frame of reference of the work. As we
shall see, as soon as we move (say) a painting from one room to
another, we have changed the frame of reference of the work itself.
But this is accidental and not related to the work itself (see ch. 22).
If a center, and by extension a format, “is the locus of
forces issuing from it and converging toward it,” the forces are
constituted of vectors. And if a center of energy is the prototype of
a centric system, then an individual vector is the prototype of an
31

32
eccentric system of dependence. Vectors are therefore “forces
generated by the shapes and configurations of perceptual objects”
(1988, p. 229, slightly amended).
The power of Fig. 3, above, depends only on its strength as
a center of energy and develops purely centrifugal behavior. “ A
different situation comes about when a second object is introduced
into the neighborhood of the first…the original center responds to
the presence of another one, a centric orientation changes into an
eccentric one. The primary centric system is no longer alone in the
world; it acknowledges the existence of other centers by acting
upon them and being acted upon by them” (1988, p. 5). This is
represented in Fig. 4.
Fig. 4. Centers reacting to centers, the basis of the eccentric com-
positional system
The relationship between the two systems, or more specifi-
cally, the interplay between centrically oriented systems and eccen-
tric vectors is not mutually exclusive. Arnheim (1988) says “if we
distinguish perceptual objects from one another by calling them
volumes or vectors, we are adopting a convenient simplification, to
be handled with caution” (p. 150). As we saw before in the discus-
sion of weight, the identification of exactly is a center determines
how weight will be determined.
The complexity of the interaction of object and action is
suggested by Arnheim in the following way:
In Fig. [5] we see the spatial order to which the composi-
tional forces conform, whereas Fig. [6] schematizes the be-
havior of these compositional forces themselves within the
given framework. In both cases the combination of two
rather disparate patterns makes the relationship quite intri-
cate. In the framework of spatial order it allows for the si-
multaneous presence of focused and homogeneous space

33
but complicates the order by creating a tricky relation be-
tween curves and straight lines. In the framework of vecto-
rial dynamics it produces the tension and discord needed
by the artist when he represents self-centered behavior as
trying for a modus-vivendi with outward-directed behavior
(p. 9).
Fig. 5. Superimposition of the two compositional systems
All works of art are subject to one particular perceptual
distortion that is similarly highly symbolic, the anisotropy of space.
Arnheim defines anisotropism as “the asymmetry of gravitational
space, by which the nature and behavior of perceptual objects
change with their location and the direction of the forces they emit
and receive” (p. 225). Perceptual anisotropy has been known since
Mach and has been investigated by gestalt psychologists as an ex-
ample of a prägnant, or singular, orientation of percepts (Koffka,
1935; Rock, 1973). We can see that the earth is the one perceptual
center relative to which it is most difficult to consider the weight of
objects. In this sense, while we can affirm that the way of looking at
a center-as-base is alternate, the earth as a center usually ‘wins out.’
All works of art share in the symbolism of anisotropy.
Symbolically, moving upward involves the overcoming of weight, a
liberation, from the ground; moving downward is experienced as
giving in to the gravitational attraction, a passive letting go (Arn-
heim, 1974, p. 30-3). This cannot be underestimated in the under-
standing of all the arts, including music. The vertical represents the

34
dimension of contemplation, a concept which we shall have the
opportunity to clarify in the next chapter. Here, it may suffice to
say that up and down relations tend to preserve the relation between
subject and object, without any getting “entangled with the vecto-
rial configuration within the plane of the configuration itself”
(1988, p. 38).
Fig. 6. The behavior of forces within a format
There are further asymmetries in the perceptual field, in
particular the right-left asymmetry, whereby the left and right sides
appear to be of different weights. Arnheim (1988) says that, “The
left side is endowed with special weight; it assumes the function of a
strong center with which the experiencer tends to identify…The left
side is also a hub, where more weight can be tolerated” (p. 47; also
Arnheim, 1974, p. 33-6). Consequently, if the vertical represents
the axis of contemplation, the horizontal represents the dimension

35
of interaction, time and narrative (Arnheim, 1977, p. 54). And so
we have the foundation of graphic meaning.
Departing from Wertheimer’s laws of perceptual grouping,
Gestaltists affirm the universality of their principles for sensory
explanation. In this sense, the schematism provided by Arnheim in
The Power of the Center suggests itself as a powerful reformulation
of Wertheimer and a basis for a general Gestalt psychology of art.

37
CHAPTER 4
THE NOTORIOUS GESTALT BRAIN MODEL
Over fifty years ago, in 1949, Rudolf Arnheim published a remark-
able article in the Psychological Review entitled 'The Gestalt The-
ory of Expression' (Arnheim, 1966). Basing his observations on the
physiological theory of his teacher, Wolfgang Köhler (1940;
Köhler & Wallach, 1944), Arnheim attempted to explain the ex-
pressiveness of visual forms as the psychological or phenomenal
counterpart to the 'stresses' and 'strains' of the underlying visual
processes themselves. For example, given a bowing column before
us, the illusion of 'bowing' is due to the compromise between a
minimization process which seeks to reduce the column to a
straight line, and the proximal stimulation which records the slight
entasis at the center of the column; the compromise survives in our
visual percept.
Arnheim went on to elaborate his ideas in Art and Visual
Perception (1954/1974) where he reiterated that the visual expres-
siveness of forms could be attributed to some kind of field-like
process occurring in the visual cortex. Later, anticipating some
developments in non-linear dynamics, Arnheim in Entropy and Art
(1971) saw perception as a compromise between minimization
('catabolic') processes and proximal stimulating ('anabolic') proc-
esses.
Arnheim's model has suffered with the fate of the gestalt
theory of brain functioning. By most estimations, this theory was
fatally crippled in the 'fifties by counterdemonstrations by Karl
Lashley, Roger Sperry and Karl Pribram (Lashley, Chow &
Semmes, 1951; Sperry, Miner & Myers, 1955; Pribram, 1971,
1984). Köhler had reasoned that the cortex operates as an electric
field and his various critics supposed that he had refuted Köhler
when they found that cats or monkeys could still perceive after
conductive metal foil, needles and metal-based creams had been
inserted into their brains – an operation that ought to have dis-
rupted any electric field. Even though gestalt psychologists have
insisted that the counter-demonstrations were not decisive (Henle,
1984), it is significant that Köhler's theory has not been supported
in toto.
Still, the advantages of Arnheim's theory should be appreci-
ated. He took a hazy subject – the expressiveness of visual forms –
and provided an empirical model. This is in sharp distinction to the
general trend in such studies where the best that is hoped for is the
elicitation of verbal judgments of the expressiveness of forms which
are then analyzed with powerful statistical instruments like the se-
mantic differential (Osgood, Suci & Tannenbaum, 1957). In the

38
following I want to point out ways in which Arnheim's intuitions
might be rehabilitated, in the same way in which gestalt brain mod-
els are no longer being simply dismissed and are receiving a second
look.
The Gestalt Brain Model: Isomorphism
The Gestalt Brain Model is based on ‘isomorphism,’ which sug-
gests that neural processes and perceptual experiences share some
common form or structure (Köhler, 1940; Köhler & Wallach, 1944;
Scheerer, 1994). This common form is only intended to be topo-
logical or functional, and Köhler rejected the so-called identity
theory of brain functioning that says that perceptual states just are
brain activities. This is often misunderstood about Gestalt theory,
and Arnheim’s writings must be framed in this particular way.
Arnheim (1949/1966) actually sought to extend this state of
affairs to complete the picture and suggest how we might extend
levels of isomorphic structure through intermediary levels:
A. Observed Person
I. State of mind psychological
II. Neural correlate of I electrochemical
III. Muscular forces mechanical
IV. Kinesthetic correlate of III psychological
V. Shape and movement of body geometrical
B. Observer
VI. Retinal image of V geometrical
VII. Cortical projection of VI electrochemical
VIII. Perceptual correlate of VII psychological
Since Arnheim was most interested in the isomorphism between
expressive contents, it led him to look beyond the brain-
consciousness isomorphism. Thus interesting is the fact that Arn-
heim includes the geometrical level at which isomorphism must be
communicated between two people communicating. This leaves
open the possibility of stimulus gradients that can communicate
expressive contents in the manner of Gibson.
Gestaltists like Arnheim attempt the difficult task of privi-
leging neither the phenomenal nor the physical, so they are neither
physicalists nor phenomenalists (Epstein & Hatfield, 1994). Phe-
nomenological experience is given its due, but at the same time so
is the transcendent brain functioning that we learn about through
science. Since Gestaltists refuse to limit reality to lived experience,
but see it critically interacting with a transcendent reality, they are
critical realists (Bischof, 1966; Mandelbaum, 1964). Thus we have

39
to distinguish between our phenomenal self and our transcendent
organism (Fig. 7; Köhler, 1938).
Fig. 7. The Critical Realist System of Epistemology, after Bischof
(1966)
For this reason Arnheim (1994/1996, pp. 144-50) calls
consciousness ‘an island of images.’ Such distinctions are not
abstract but point to the difficulties of psychological explanation of
the arts. The philosopher Monroe Beardsley (1979) has discussed
the problem of ‘objective’ and ‘psychological’ language in Arn-
heim’s theories and complained they become intertwined in his
writing. Arnheim would answer that we need both, used properly
when speaking of the transcendent world or the phenomenological
world.
We can also say that perceiving is not an isolated activity for
in our very acts of perceiving we make causal inferences about the
world that are quasi-scientific. Thus we make phenomenal observa-
tions about the arts (or the world) that guide hypotheses about un-
derlying processes, and these lead to observations. There is no point
of innocence.
Keeping the Field Metaphor, Jettisoning the Specific Mechanisms
Naturally, Köhler made several assumptions about the brain in the
'fifties which underestimated its complexity and his ideas about
simple fields of action, although suggestive, cannot be supported at
the level of proposed mechanisms. At the same time, the emerging
science of point recordings of the action of feature detectors –
picking up sensitivity to a bar, texture or grating – was lamented by

both Köhler and Arnheim (1971). Contemporary accounts of mi-
croscopic dendritic arborization and feature detectors – the so-
called ‘neuron doctrine’ – still cannot account for the complexity
and richness of human perception, suggesting that some more
global perspective is still needed. Spillmann & Ehrenstein (2004)
point out that the old, mechanistic picture of single cells has been
overcome by a much more impressive field-like ability to capture
relational data. These in turn suggest larger coordination, leading to
the concept of ‘perceptive neurons’ (Baumgartner, 1990) for cells
that respond as though they mediated the presence of illusory
contours (with neither a physical, nor retinal stimulus correlate).
What must be kept is the meta-theory. Arnheim partially
did this in his little work, Entropy and Art (1971), in which this
psychologist courageously took on notions of entropy to argue that
physics had yet to develop tools to deal with the ‘upward’ character
of physical systems. At this very time, cybernetics and systems the-
ory was just giving way to the first discoveries of synergetics and
catastrophy theory.
Two recent interpreters of the gestalt tradition have done
precisely this. The first was Arnheim’s colleague, Erich Goldmeier
(1982), who obtained his doctorate under Wertheimer in the 'thirties
in Frankfurt. In the context of a theory of memory, Goldmeier
likened 'prägnant' or prototypical forms to minimization wells in a
hypothetical brain process. Goldmeier, who was also trained in
physics, had catastrophe theory in mind when he made his analo-
gies.
Around the same time, Michael Stadler in Germany was de-
veloping a brain theory based upon the principles of synergetics of
the physicist Hermann Haken, with whom he collaborated. Stadler
was an ideal candidate to continue Köhler's approach and had writ-
ten on after-effects based on Köhler's model as early as the 'sixties.
Like Goldmeier, Stadler agreed that prototypical forms could be
likened to potential wells in a hypothetical space or else to syner-
getic 'attractors' governing the process. Stadler's approach has been
slightly more dynamic than Goldmeier's, allowing for gestalten to
emerge spontaneously from processes in which synergetic proc-
esses are at work.
Both Goldmeier and Stadler believe that the spirit of
Köhler's theory was correct, even if the specific mechanisms he
proposed are not. As Stadler and Peter Kruse (1990) have written,
“the apparent incompatibility of the analytical and the holistic
system view seems to be more a fruitful starting point than an em-
barrassment” (p. 35). In the same vein, Steven Lehar has argued
that what is required of neurobiological speculation is ‘perceptual
modeling,’ that is, seeking hypothetical physical processes and
mathematical formalizations adequate to capture phenomenal data.
40

Arnheim, Gestalt and Art A Psychological Theory.pdf

Arnheim, Gestalt and Art A Psychological Theory.pdf

Recommended

Recommended

More Related Content

Similar to Arnheim, Gestalt and Art A Psychological Theory.pdf

Similar to Arnheim, Gestalt and Art A Psychological Theory.pdf (20)

More from Samantha Martinez

More from Samantha Martinez (20)

Recently uploaded

Recently uploaded (20)

Arnheim, Gestalt and Art A Psychological Theory.pdf